Hacker News Reader

Comment by smcl | Hacker News Reader

It's amazing to think how far we've come since June 2018 where the record was just shy of three minutes (https://www.youtube.com/watch?v=TLbfqZBL8t8). These new strategies could allow experienced speedrunners to exit vim in times of 10 seconds (or perhaps less!?). I'm excited to see how this progresses, the vim speedrun community is vibrant and creative and I've no doubt we'll see a sub 5-second run within my lifetime.

View on HN

I'm confident we're at the dawn of a new era of AI driven vim closing.

My first attempts ended up with my bot buying stock in a clorox brand before posting questionably racist comments to my twitter, but these are minor concerns on the journey.

I'm streaming the process on twitch but please note that donations won't be considered as seed capital for stock assignment

I believe the solution to this lies in using Rust and SIMD. It will achieve both performance and safety.

I would prefer a functional approach and formal verification so we can know for sure it will exit.

Since the real world is not necessarily deterministic as demonstrated by the methods using random(), I hope to see more development in stochastical methods describing parameters of probability distributions of successful exits.

In the face of vim's unyielding resilience to exiting, I think we may need to devise a new algorithm for this using quantum superposition. I heard IBM can help with that.

But who would support it? While a formally verified functional approach may satisfy one's doctorate requirements, the author will graduate and move onto blockchain development, leaving behind an exit-vim code base that no other engineer could possibly understand without the equivalent training. Upon receiving said training, an engineer would also take newly acquired doctorate and also move onto blockchain development. Essentially, supporting a formally verified functional approach to exiting Vim creates a pipeline of well-credentialed and rigorously trained professionals for blockchain development. If Vim were, however, put on the blockchain, we may be able to address this concern and you can get what you wanted.

Or, you could have just used Rust and SIMD.

This hits so close to home.

Let me halt you right there...

There is a much faster way, but first you need to build up speed for 12 hours.

Pretty sure there will be a TAS in the sub-second range in the near future!

Must be GDQ season again.

I'm curious if there is any work in fuzzing that could be applied here.

Reply on news.ycombinator.com